Thread

1. Bug in parser?

   Gerhard Dieringer <gerhard.dieringer@arcor.de> — 2003-05-31T20:16:24Z

   
   ============================================================================
                           POSTGRESQL BUG REPORT TEMPLATE
   ============================================================================
   
   
   Your name		:	Gerhard Dieringer
   Your email address	:	Gerhard.Dieringer@nexgo.de
   
   
   System Configuration
   ---------------------
     Architecture (example: Intel Pentium)  	:Dual Intel Pentium II
   
     Operating System (example: Linux 2.0.26 ELF) 	: Linux 2.2.16
   
     PostgreSQL version (example: PostgreSQL-7.3.3):   PostgreSQL-7.3.3
   
     Compiler used (example:  gcc 2.95.2)		:   gcc-Version 3.3
   
   
   Please enter a FULL description of your problem:
   ------------------------------------------------
   Inconsistent results when calling '+' operator with text arguemts
   The last 4 results (f.e. '1' + '2' -> 'c') are very strange.
   
   
   
   Please describe a way to repeat the problem.   Please try to provide a
   concise reproducible example, if at all possible: 
   ----------------------------------------------------------------------
   select 1 + '2' as result;
    result 
   --------
         3
   (1 row)
   
   select 1 + '2.3' as result;
   psql:error.sql:2: ERROR:  pg_atoi: error in "2.3": can't parse ".3"
   select 1.5 + '2' as result;
    result 
   --------
       3.5
   (1 row)
   
   select 1.4 + '2.3' as result;
    result 
   --------
       3.7
   (1 row)
   
   select '1' + '2' as result;
    result 
   --------
    c
   (1 row)
   
   select '1' + '2.3' as result;
    result 
   --------
    c
   (1 row)
   
   select '1.5' + '2' as result;
    result 
   --------
    c
   (1 row)
   
   select '1.4' + '2.3' as result;
    result 
   --------
    c
   (1 row)
   
   
   
   
   
   If you know how this problem might be fixed, list the solution below:
   ---------------------------------------------------------------------
   Sorry :-((
   
   
   
   
   

2. Re: Bug in parser?

   Stephan Szabo <sszabo@megazone23.bigpanda.com> — 2003-06-01T15:05:36Z

   On 31 May 2003, Gerhard Dieringer wrote:
   
   >
   >
   > ============================================================================
   >                         POSTGRESQL BUG REPORT TEMPLATE
   > ============================================================================
   >
   >
   > Your name		:	Gerhard Dieringer
   > Your email address	:	Gerhard.Dieringer@nexgo.de
   >
   >
   > System Configuration
   > ---------------------
   >   Architecture (example: Intel Pentium)  	:Dual Intel Pentium II
   >
   >   Operating System (example: Linux 2.0.26 ELF) 	: Linux 2.2.16
   >
   >   PostgreSQL version (example: PostgreSQL-7.3.3):   PostgreSQL-7.3.3
   >
   >   Compiler used (example:  gcc 2.95.2)		:   gcc-Version 3.3
   >
   >
   > Please enter a FULL description of your problem:
   > ------------------------------------------------
   > Inconsistent results when calling '+' operator with text arguemts
   > The last 4 results (f.e. '1' + '2' -> 'c') are very strange.
   
   '1'+'2' is pretty meaningless.  In the other cases you're giving a type
   that has a meaningful + operator so it's trying to convert  the quoted
   argument to an appropriate type for +.  In the '1'+'2' case it's
   converting to "char" (single character) and adding those which seems
   marginally reasonable to me (although I think that "char" is pretty
   silly).
   
   I'm not sure what you expected '1'+'2' to give though, concatenation is ||
   and I can't think of something meaningful to do with it.
   
   
   

3. Re: Bug in parser?

   Tom Lane <tgl@sss.pgh.pa.us> — 2003-06-01T15:32:26Z

   Gerhard Dieringer <Gerhard.Dieringer@arcor.de> writes:
   > Inconsistent results when calling '+' operator with text arguemts
   
   I suspect you are looking for the '||' operator, which is the SQL-standard
   spelling of concatenation.  '+' ends up invoking the one-byte-"char"
   datatype's addition operator.
   
   			regards, tom lane